You are here

Sync Progress & some Suggestions

6 posts / 0 new
Last post
nokPT
Offline
Last seen: 6 months 1 day ago
Joined: 2009-12-18 12:22
Sync Progress & some Suggestions

I usually use SyncToy, but I'm having some errors and some fails in sync.

I started a sync between my company server and a portable hard disk.

My sync job is some what big: 26.379 Files, 2.576 Directories, 14,9GB

I’ve configured then sync job and made Preview, after a few minutes, it updated the destination tree

=> Suggestion: while in Preview, it should show some activity, SyncToy shows the path of the file.

After Preview, I pressed Run, a new window come up, saying:
“17:22:18 Starting”
“ Setting up progress bar”
It remain like this until about 17:42, it took 20 minutes to show some action (copy/remove/etc…), if copied for about 9 minutes, after another big pause (except the progress bar that some times added a new bar – it was working), 18:11 copied 2 new files, 18:28 removed a lot of files, 19:35 to 19:37 copied and removed some files (very slow, it took some seconds to delete small files in the same directory)...

=> Suggestion: Like my upper suggestion, it should always show some thing, if the file is identically, it should log it in the Progress Window, this way we know it is working, and in what point.

=> Suggestion: In the run window, every line should/could have a colour, just link in the tree after Preview, it is better to see.

=> Suggestion: In big files, it would be nice to say, starting copy… copied, in SyncToy, the progress bar shows the progress of this file, not all job.

=> Another Suggestion: On the top, on the right of the Run, Preview and refresh buttons, you could place a legend with the colour and what it means

I’m not saying that SyncToy is better, only that its feed-back helps more and also it looks faster, actualy a lot faster.

Is the any thing we can do to improve your program?

It is now 18:40, it’s about 20% of the 1st job done (one and half hour and just 20%), I think it is to slow, and that a simple delete *.* and xcopy *.* /S should be faster.

Realy: Is the any thing we can do to improve your program?

EDIT: 21-12-2009, It took 4 hours and 14 minutes to do 1st job, I have 3 jobs, it is very slow, I'm using a 100MB Ethernet, and at the moment only 3 computers are connected to the network

Steve Lamerton
Steve Lamerton's picture
Offline
Last seen: 9 years 2 months ago
Developer
Joined: 2005-12-10 15:22
I

think the main problem here is the difference between what is actually happening as what Toucan shows. Because of the way Toucan works it only copies files that have actually changed, the idea behind this is to reduce the amount of writes that Toucan performs as it is frequently used on flash drives with limited write spans. Currently when Toucan skips a file nothing is output and the progress bar is incremented, however as you noticed with a lot of files this is very hard to see, it will be rectified as of the next version to give better feedback as you have suggested.

As for speed this is a known weakness with Toucan, especially over networks where its read-heavy approach is limited, I intended to add auto-detection of the type of connection it is working over so it can optimise for various situations.

Thanks for the feedback Smile

LucheLibre
Offline
Last seen: 9 years 4 months ago
Joined: 2009-10-27 16:24
You're not kidding....

You're not kidding about the read-heavy nature of Toucan. I'm really surprised at its methods.

I did a Process Monitor trace of Toucan while it was syncing to a whs share, and I was amazed at how Toucan will read each file simulaneously 4KB at a time (sometimes repeating the same block a time to two), resetting IO flags with each read.

While probably being very reliable and thorough, I've got to imagine that a better use of buffers or some other comparison technique would help a lot.

I started a sync job in which 99% of the two volumes were identical. Despite that, Toucan reads each and every file in its usual way. As an example, the comparison of one 2.5M JPEG between my laptop and my server (just reads, no writes) involved nearly two thousand read operations and took 3 seconds. That is a crazy-slow and inefficient rate.

(Update: Read the "MD5" thread and am looking forward to 3.0 and the new comparison method. Going to check out the streaming toggle, too. It will probably cure my ills. Sorry for the rant, Mr. Dev.)

"Save us, Porta-Jesus!"

(because "Porta-John" was already taken)

nokPT
Offline
Last seen: 6 months 1 day ago
Joined: 2009-12-18 12:22
Worst than that -> I is not reliable

I’ve being using it, on weekly bases, to an alternative backup storage (so I used alternative backup software -> toucan), some times it took 2 hours other times 4 hours.

Last 2 times, I had to interrupt the backup procedure.

Don’t know why, but today I’ve decided to compare the Mirror time with the Echo time of the SyncToy (give a second chance to toucan), so first I made a preview (in synctoy) and it took a lot of time, don’t know if less or more than toucan, but it isn’t important because, synctoy detected over 21.000 files not copied, most of them were there from the start.

This is very bad, If it is slow -> we wait, if it missies files -> we don’t use.

I don’t know why missed so many files, because it doesn’t make any sense, it missed directories completely, I’m from Portugal, and we have special characters like: ç or ã or õ or â or …, but it copied some files and directories with that characters, so this isn’t the problem.

Thank Gog I’ve tried another software or I could have a big problem.

Best of luck

Steve Lamerton
Steve Lamerton's picture
Offline
Last seen: 9 years 2 months ago
Developer
Joined: 2005-12-10 15:22
I

am not aware of there being any issues like this in the current version of Toucan, it certainly has full support for the characters you mention. Is there any way you can narrow down what you think it missed? Thanks.

nokPT
Offline
Last seen: 6 months 1 day ago
Joined: 2009-12-18 12:22
At this point I can't narrow

At this point I can't narrow down, but it missed a lot, missed more than 8.000 files in a total of over 26.000 files.

I remember this one
In one of my directories that is:
Source -> "\\Central\Projectos Automação\Manuais"
Destination -> "H:\Backup\CSC\Projectos Automação\Manuais"

In this directory I have about 10 files plus about 20 sub-directories with a lot of files, mostly pdf files, it just copied 4 sub-directories, it missed all the files (of that directory) and all other sub-directories.

I've looked for a log in Toucan directory but there isn't any, so bump.

Other thing, about the time takes to do the job: With Synctoy:
* it took about 10 minutes to preview, I don't know how many time Toucan took
* it took about 20 minutes to mirror all files and the was a big file with 1 Giga (or about 2 or 3 minutes without that 1 Giga file), with Toucan the best (minimum) time was over 2 hours (an it missed several times a lot of files)

If I have the time, I'll try to make a new job to another disk (or some thing) and try to capture some thing, or make a list of missed files, any thing that can help.

Log in or register to post comments